Scribbler Logo

    You have 4 summaries left

    DataFramed

    #140 How this Accenture CDO is Navigating the AI Revolution

    Mon Jun 05 2023
    Data TransformationGenerative AILife SciencesData BacklogClinical TrialsAI ImpactData PrivacyTeam DiversityHealthcare Advocacy

    Description

    This episode covers a wide range of topics related to data transformation, generative AI, and its applications in life sciences. It explores the challenges of data backlog, the importance of data literacy, and the need for responsible AI practices. The episode also delves into model-informed drug discovery, clinical trial adherence, and the role of AI in marketing materials. It discusses the impact of AI on data processes, the balance between data privacy and democratization, and the significance of team diversity. The episode concludes by emphasizing the importance of advocating for our own healthcare and embracing the potential of generative AI.

    Insights

    Data transformation is essential for organizations

    Data transformation involves modernizing technical architecture, resetting the value of data, and embracing a fundamentally different approach.

    Generative AI can automate repetitive tasks

    Generative AI allows humans to focus on more engaging work by automating repetitive tasks.

    AI can accelerate drug discovery processes

    AI has the potential to streamline product pipelines and accelerate drug discovery processes in life sciences.

    Data privacy and security are crucial in life sciences

    Data privacy and security are paramount in life sciences applications of AI, requiring careful curation and protection of health data.

    Building diverse teams is important for responsible development

    Diversity in teams is crucial for challenging the status quo and ensuring responsible development of AI capabilities.

    Advocating for our own healthcare is essential

    In the data and AI-driven life sciences industry, it is important to advocate for our own healthcare and be aware of advancements and opportunities.

    Chapters

    1. Introduction
    2. Data Backlog and Transformation
    3. Model-Informed Drug Discovery and Clinical Trials
    4. Data Processes and Generative AI
    5. AI in Life Sciences and Data Privacy
    6. Data Privacy, AI Impact, and Team Diversity
    7. Building Diverse Teams and Advocating for Healthcare
    8. Conclusion
    Summary
    Transcript

    Introduction

    00:00 - 07:19

    • Large language models in the root of chat GPT is not new. We've been doing this since the 50s, right? The difference is we're able to paralyze it. We're able to check sentiment. We're able to do it at a speed, this meaningful. That's what's changed. And then of course, the final tranche of this, which is that it is totally democratized and that it's in everyone's hands.
    • Tracy Brink is the Life Sciences Chief Data and Analytics Officer and Global Generative AI Lead at Accenture.
    • Data modernization focuses on modernizing the technical architecture, while data transformation resets the value of data and how it's leveraged in an organization.
    • Data transformation involves thinking about data in a fundamentally different way and working with business stakeholders to change its embrace.
    • Data literacy or fluency plays a crucial role in data transformation programs.
    • Business end users, source systems, feeder systems, and focus on change management are all involved in data transformation programs.

    Data Backlog and Transformation

    06:49 - 13:56

    • A company had a two-year backlog of data work that needed to be done.
    • 30% of their pipeline was duplicative and 10% was already available but unknown.
    • The backlog was shrunk down and caught up in four months.
    • Creating literacy, transparency, and communication is important for solving data backlog issues.
    • Many organizations have multiple dashboards for every employee worldwide, which is inconsumable and creates technical debt.
    • Keeping technological and data house tidy is crucial.
    • Data transformation programs in life sciences require a North Star metric to guide the journey.
    • Agility and real-time analytics are important factors in decision-making for commercial or clinical trials.
    • Responsible AI, data sharing, privacy, and security are common challenges across industries including life sciences.
    • Life sciences can learn from other industries like consumer products in terms of customer experience.
    • AI can be used to streamline product pipelines and accelerate drug discovery processes.

    Model-Informed Drug Discovery and Clinical Trials

    13:39 - 20:35

    • Model-informed drug discovery can help accelerate the process of bringing drugs to market and into clinical trials.
    • Dropping out early in the drug discovery process can save time and resources.
    • Maintaining and utilizing data from dropped-out indications can be valuable for future research.
    • Clinical trial adherence is crucial, and new ways of communication and nudging can improve it.
    • Using weather patterns to inform patients about potential disruptions to their clinical trial appointments can help ensure compliance.
    • Marketing materials need to be consistent and compliant with regulatory agencies, while also being tailored to specific doctors' preferences.
    • Generative AI can automate repetitive tasks, allowing humans to focus on more engaging work.
    • Managing different systems requires careful consideration of data integration and responsible AI practices.
    • Processes around data need to be revisited when using AI, as we are in uncharted waters.

    Data Processes and Generative AI

    20:16 - 27:18

    • Processes around data need to change when using AI.
    • All processes need to be revisited in the era of AI.
    • Data DevOps is the standard now, replacing traditional release schedules.
    • Clients may not have the budget or interest for process reinvention, so new processes should be implemented alongside value creation.
    • Implementing new processes and procedures while bringing new value is a best practice.
    • Prioritizing generative AI depends on the state of existing data pipelines and technological debt.
    • Generative AI can expose technological debt at a faster pace than before.
    • Finding a targeted use case and gaining traction is a common approach to starting with generative AI.
    • Cleaning up existing systems is necessary before fully embracing generative AI.
    • Life Sciences companies are adopting digital core or digital fabric approaches for agility and responsiveness.
    • Digital core enables more agile responses to future challenges.
    • Examples of AI use cases include automation in content management, regulatory filings, compliance in life sciences and pharma, and intelligent supply chain optimization.
    • AI can help prevent equipment failures and financial losses in specialty medication production.

    AI in Life Sciences and Data Privacy

    26:51 - 33:56

    • Predictive maintenance using AI helped catch a potential failure in a client's system
    • AI can build trust when it consistently delivers accurate results
    • Using generative AI for regulatory filings requires careful consideration of accuracy and human confirmation
    • AI can help catch mistakes in regulatory filings by training models on historical data
    • Data privacy and security are paramount in life sciences applications of AI
    • Health data needs to be carefully curated and protected
    • Democratizing technology is important, but not at the expense of data privacy
    • CDOs grapple with the balance between protecting and democratizing data

    Data Privacy, AI Impact, and Team Diversity

    33:28 - 40:41

    • CDOs often grapple with the decision of whether to hold and protect data or democratize it
    • Data privacy and security are of material importance, especially at the board level
    • AI is becoming an integral part of various systems, including CRM, ERP, and data governance
    • Measuring the impact of AI is still a learning process, but efficacy, safety, and accuracy are key metrics in life sciences
    • Building an AI capability requires hiring staff with AI fluency across the organization
    • Diversity in teams is crucial for challenging the status quo and ensuring responsible development

    Building Diverse Teams and Advocating for Healthcare

    40:13 - 46:56

    • Recruiting focuses on core technical capabilities and building a diverse team that challenges the status quo.
    • WLDA (Women Leaders in Data and AI) is an invitation-only group supporting women and allies in data and AI careers.
    • Tech fluency is important, and peer-based learning and networking are valuable for growth.
    • Creating a diverse team can be practical by adding new talent and avoiding comfortable familiarity.
    • Being watchful of bias in decision-making processes is crucial for effective teamwork.
    • Continuing to work together, improving meetings, and fostering infinite curiosity are key to success.
    • Advocating for our own healthcare is more important than ever in the data and AI-driven life sciences industry.

    Conclusion

    46:36 - 48:15

    • Being our own advocates for healthcare is more important than ever.
    • Impressed with the advancements in the healthcare industry and hopeful for better care, experiences, and outcomes.
    • Data and AI are improving every day, benefiting doctors and caretakers.
    • Encourages audience to explore generative AI and continue experimenting.
    1